FOG Multicast Problems / Partclone fails to finish

[Joe Gill]

I have FOG server 1.2.0 currently installed on an Ubuntu 14.04 Linux desktop Virtual Machine and a FOG Storage Node installed on a another Ubuntu 14.04 Linux desktop also a Virtual Machine. We have a Windows Domain Controller and a Windows DNS server one running Windows Server 2008 and the other Windows Server 2012. We are a school and all of our clients are either Windows 10, Windows 7 (mostly) and some Windows XP.

I have had many hurdles in installing and running FOG. Currently, I have finally gotten Unicast figured out and running. However, I cannot get multicast to run. Partclone gets to the start screen and then will not finish.

I have gone through this wiki:
https://wiki.fogproject.org/wiki/index.php?title=Troubleshooting_a_multicast

But cannot figure out how to:
“Now boot up 1 client go to your FOG Menu and select debug mode.”

The server starts udp-sender just fine with no errors. I can also log in and move my files around in FTP on all of my storage nodes. My FOG Storage is the default storage node.

My FOG Multicast Settings are:

FOG_UDPCAST_INTERFACE = eth0 (which is correct)
FOG_UDPCAST_STARTINGPORT = 58166
FOG_MULTICAST_MAX_SESSIONS = 64

It sounds like some of these problems have been resolved in TRUNK but I cannot figure out for the life of me how to install the TRUNK version in UBUNTU 14.04.

I know my network settings are correct because I can use Clonezilla multicast just fine.

Any help is greatly appreciated.

I’m a longtime linux/unix user but it’s been a while, so please dumb down the replys a bit. Thanks!!

Cheers,

Joe Gill

[Joe Gill]

@Joe-Gill

I did mange to download the latest trunk version using svn. I got it downloaded and am going to install it and see what happens. In the meantime help is appreciated. Thanks!!

Cheers,

Joe Gill

[Wayne Workman]

when it refers to “debug”, you would just do a multicast deployment of said image to said machine, and on the confirmation window in FOG Trunk, tick the “debug” checkbox. This will cause said host to boot to a shell.

[Sebastian Roth]

@Joe-Gill said in FOG Multicast Problems / Partclone fails to finish:

I know my network settings are correct because I can use Clonezilla multicast just fine.

While this sounds like your network is setup correctly can you still give us some more details about what and how you tested with clonezilla.

As well you want to check the log (/var/log/fog/multicast.log) and FOG services running (ps aux | grep FOGMulti).

[Joe Gill]

Thanks for the responses!

I was able to download the latest trunk version. Yay! So now I’m up to date on my install.

I was also able to run the debug on multicast. I can successfully run a multicast with two clients in debug mode. So it must be a setting in FOG.

As for Clonezilla… I used it to image one of our labs. Basically I put the server edition on a USB flash drive, booted to it and pushed my images from their off of our NAS. This worked well. However, it does not provide all of the functionality of FOG.

Also, I did check the /var/log/fog/multicast.log file out. The majority of the errors there say:

This is not the master node.

I checked to make sure I have our primary node set correctly in the GUI and it is correct. I have the FOG storage as Master Node.

The only thing else that’s strange is that it says:

Interface Ready with IP Address: 65.XXX.XX.194 address listed which is our public IP.

Any ideas are appreciated!

Thanks!!

Cheers,

Joe Gill

[Joe Gill]

Here is the results from the ps command… It appears the service is running.

root@Fog:/# ps aux | grep FOGMulti
root 25118 0.0 0.2 203568 24940 ? S 12:43 0:00 /usr/bin/php -q /opt/fog/service/FOGMulticastManager/FOGMulticastManager
root 25121 0.0 0.1 283860 19628 ? S 12:43 0:00 /usr/bin/php -q /opt/fog/service/FOGMulticastManager/FOGMulticastManager
root 26380 0.0 0.0 15948 2244 pts/7 S+ 14:08 0:00 grep --color=auto FOGMulti

[Tom Elliott]

@Joe-Gill It appears twice though, both with different parent pids?

Can you run:

service FOGMulticastManager stop; sleep 10

Post output of ps aux|grep FOGMulti again? Yes I nkow I’m asking you to stop the service, but I need to see if it thinks there’s two services running at the same time.

[Tom Elliott]

Sorry I’m an idiot.

Please do output of ps -ef | grep 'FOGMulti' | grep -v grep

[Joe Gill]

Hi Tom,

Here is the result…

root@Fog:/# ps -ef | grep ‘FOGMulti’ | grep -v grep
root 25118 1 0 12:43 ? 00:00:00 /usr/bin/php -q /opt/fog/service/FOGMulticastManager/FOGMulticastManager
root 25121 25118 0 12:43 ? 00:00:01 /usr/bin/php -q /opt/fog/service/FOGMulticastManager/FOGMulticastManager

Joe

[Tom Elliott]

@Joe-Gill Now this particular node IS a master node?

[Joe Gill]

This was after I listened to you and ran the first command…

root@Fog:/# service FOGMulticastManager stop; sleep 10

• Stopping FOG Computer Imaging Solution: FOGMulticastManager [ OK ]
  root@Fog:/# ps -ef | grep ‘FOGMulti’ | grep -v grep

As you can see nothing is running now.

[Tom Elliott]

@Joe-Gill Good, now start it back running with service FOGMulticastManager start

[Joe Gill]

@Tom-Elliott

ps -ef | grep ‘FOGMulti’ | grep -v grep

Revealed nothing running.

Should I try re-running the multicast?

[Joe Gill]

And yes this is the master node. I removed my secondary storage node just to be sure. I’ll go check the logs since I did that.

Restarted a Multicast with 2 machines and it’s stuck at the Partclone screen.

[Joe Gill]

I checked the /var/log/fog/multicast.log, and it still says:

[06-16-16 8:37:57 pm] Interface Ready with IP Address: 65.XXX.XX.194
[06-16-16 8:37:57 pm] Interface Ready with IP Address: 172.16.1.17
[06-16-16 8:37:57 pm] Interface Ready with IP Address: Fog
[06-16-16 8:37:57 pm] * Starting MulticastManager Service
[06-16-16 8:37:57 pm] * Checking for new items every 10 seconds
[06-16-16 8:37:57 pm] * Starting service loop
[06-16-16 8:37:57 pm] | This is not the master node

How do I get rid of the 65.xxx.xx.194 address? This is our outside address for our public ip address.

[Sebastian Roth]

@Joe-Gill You don’t need to get rid of the public address. This is just some kind of information but FOG wouldn’t use that interface unless you configure it to do so.

[06-16-16 8:37:57 pm] | This is not the master node

Well, that’s something I’d say. Please check your storage configuration in the web gui…

[Joe Gill]

[Joe Gill]

Also, I did add back my Freenas storage node as an additional storage node but it is not master. I can’t seem to wrap my head around this. I did try uploading a new image with the new version of partclone. I’m going to try re-pushing a multicast on Monday. But I don’t think it’s going to matter much.

The one thing that is strange in this install is that when I initially set up the server I had my Freenas storage as the only node until I had problems. Then I demoted the Freenas down to slave and installed the FOG storage node. My main image was originally stored on the Freenas server when it was the master. But I’ve since reimaged that machine several times onto the new FOG storage.

[Tom Elliott]

@Joe-Gill Where is 172.16.1.22 located on that same storage node?

Your log shows 65.x.x.x and 172.16.1.17, not 172.16.1.22 as listed on the storage node configuration.

[Joe Gill]

@Tom-Elliott

Hi Tom,

65.x.x.x is our Public IP address on our router.
172.16.1.17 is the address for our FOG Server.
172.16.1.22 is the address for our FOG Storage Node.

Thanks!

Joe